Individualising Radiotherapy Through Mechanistic Models

Radiotherapy treats cancer through the precise delivery of high doses of radiation to tumours, killing cancerous cells by damaging their DNA. Radiotherapy is highly effective because radiation can be accurately targeted to tumours and avoid normal tissue, preventing the unwanted side effects which would result from killing healthy cells. The introduction of new advanced treatment techniques and better imaging to improve tumour targeting has significantly improved patient outcomes following radiotherapy in recent years. However, while radiotherapy benefits from a high degree of physical personalisation, more can be done to improve treatment outcomes.

Cancer is a highly complex disease, and is associated with a large number of different types of mutation. These mutations can significantly affect the radiation sensitivity of a given patient's cancer. Despite this, all patients with cancer in a particular organ are typically treated with the same dose and treatment schedule. While these doses have been tailored to cancer at a population level, this almost certainly under- and over-treats some patients. If individual radiosensitivity can be precisely defined before treatment, significant improvements in outcome could be achieved, in terms of improved tumour control or reduced side effects, depending on the patient's particular genetics.

This project seeks to address this challenge by developing models of how cells respond to radiation, which can accurately predict the sensitivity of an individual's disease based on the mutations present in their particular cancer. This work seeks to answer a number of questions, including:

1. How the initial radiation interacts with the cell to cause DNA damage. This will use mathematical modelling techniques from the physical sciences to calculate how energy is deposited in individual cells, and how this causes damage to individual DNA strands.

2. How cells respond to this initial damage. Here, we will model how cells respond to different distributions of DNA damage, including how likely they are to repair this damage, and how likely the cell is to survive following a given radiation exposure.

3. How patient genetics impacts on these responses. While we know the processes which determine how sensitive a cell is to radiation (for example, its ability to repair DNA damage), it is difficult to measure these for each patient. Instead, we will develop methods to predict how effective these processes are based on the genetics of the individual's disease, which can be directly measured before treatment.

4. How clinical treatments can be optimised to incorporate this knowledge. Based on these models, we will then develop a tool which will allow for the best radiotherapy treatment to be designed taking into account both physical and biological personalisation, to maximize the chance that each patient's disease will be successfully treated with minimal side-effects.

If successful, this research programme will deliver a unique tool to enable the targeting of radiotherapy using both physical and biological factors, offering more personalised therapy and better treatment outcomes for patients suffering from cancer in the future.

In addition to the academic community, the development of effective, mechanistically-informed models of individual radiosensitivity will have impacts on a number of areas of society. These include:

1. Clinicians and Healthcare Professionals: Treatment optimisation remains a significant challenge in clinical radiotherapy, with many patients being under- and over-treated due to lack of information on individual radiosensitivity. If these novel mechanistic models of radiation sensitivity are translated into clinical tools it will open the prospect of routine radiotherapy personalisation. This may have a number of impacts on clinical practice in radiotherapy. In addition to the obvious benefit of providing more tailored doses of radiotherapy, reducing side effects or increasing control, it may also enable more rational comparisons of the trade-offs between the inclusion of potential adjuvant therapies or alternative treatment modalities such as surgery. In addition, tools for stratifying patient responses to different types of radiation may help refine treatment decisions around scarce resources such as proton therapy. These improvements will have the potential to significantly impact on the landscape of clinical radiotherapy in the future.

2. Public Sector & Policy Makers: Related to the above, quantitative models of individual response to radiotherapy may have a significant impact on policy decisions around radiotherapy provision in general. Current radiotherapy provision is based around treatments standardised on the population level, while personalised radiotherapy courses may significantly impact on this decision - both in terms of overall provision of radiotherapy, as well as prioritisation between different spending decisions. This is particularly relevant around sensitivity to proton therapy, as high capital costs associated with proton centres mean it is currently a relatively scarce resource, and is likely to remain so in the near future. Current patient selection focuses on the improved physical dose distributions with protons, but do not incorporate the potential for biological optimisation. Development of a biomarker for proton RBE may give a better insight into the true potential benefits of proton therapy, and help redefine patients who are recommended for proton therapy. These developments would help to improve the efficiency of radiotherapy delivery within the UK health service and better inform future resource allocation.

3. Patients: While important, impacts in the above sectors are primarily of interest because they will lead to significant improvements for the key end beneficiaries of this work, radiotherapy patients. The ability to deliver personalised radiotherapy is expected to reduce side-effects and improve tumour control compared to current 'one-size-fits-all' approaches. 130,000 people receive radiotherapy each year in the UK, and the vast majority of these would in principle be candidates for some degree of personalisation if suitable models were available. As a result, even modest improvements in outcome have the potential to translate into substantial improvement in survival or quality of life across the population as a whole.

4. Commercial Partners & Industry: The development of biomarkers of treatment sensitivity is a major area of activity for the UK bioscience industry. The conversion of the results of this work from a research tool to a validated clinical biomarker will necessitate close involvement with commercial partners, and may serve to stimulate further interest in developing new commercial applications of this mechanistic modelling approach. There may also be interest from companies involved in the delivery of radiotherapy planning tools in incorporating these types of models at the radiotherapy planning stage, to further refine and personalise treatment.